A platform for research: civil engineering, architecture and urbanism
Most conventional Origin-Destination models work with limited link observations with help of target demand or other prior information. This paper presents a new approach for estimation of Origin-Destination flows and route choice parameters simultaneously from incomplete link flows. By presenting approximate joint probability distribution functions of the temporal link traffic flows on a network, the conditional likelihood functions for estimating dynamic O-D and route choice parameters are derived. An Expectation-Maximization (EM) algorithm is proposed to estimate those parameters based on incomplete link flow data. Preliminary results indicate that the new approach can provide validated results.
Most conventional Origin-Destination models work with limited link observations with help of target demand or other prior information. This paper presents a new approach for estimation of Origin-Destination flows and route choice parameters simultaneously from incomplete link flows. By presenting approximate joint probability distribution functions of the temporal link traffic flows on a network, the conditional likelihood functions for estimating dynamic O-D and route choice parameters are derived. An Expectation-Maximization (EM) algorithm is proposed to estimate those parameters based on incomplete link flow data. Preliminary results indicate that the new approach can provide validated results.
Estimation of Origin-Destination Flows and Route Choice Parameters from Incomplete Link Flows
He, Rong (author)
International Workshop on Computing in Civil Engineering 2007 ; 2007 ; Pittsburgh, Pennsylvania, United States
Computing in Civil Engineering (2007) ; 347-354
2007-07-23
Conference paper
Electronic Resource
English
Estimation of Origin-Destination Flows and Route Choice Parameters from Incomplete Link Flows
British Library Conference Proceedings | 2007
|SPATIAL ECONOMETRIC MODELING OF ORIGIN-DESTINATION FLOWS
Online Contents | 2008
|British Library Online Contents | 2001
|